Using STAR-CCM+ for Catalyst Utilization Analysis

Transcription

1 Using STAR-CCM+ for Catalyst Utilization Analysis Amsterdam Netherlands March W.U. A. Leong Dunton Technical Centre Ford Motor Company S. Eroglu and S. Guryuva Gebze Engineering Ford Otosan Page 1

2 Contents Background Benefits of using CFD for Exhaust Product Development Assumptions Key Features of Current Approach Objectives of the STAR-CCM+ Upgrade Current Status Verifications Conclusions Page 2

3 Background A number of years ago, Ford Motor Company (FMC) suffered a catalyst recall in North America. To avoid such issues happened again, a CFD-based method was developed to optimise catalyst gas flow distribution. The original methodology was based on under-floor exhaust systems but the current test procedure is applicable to hot-end designs with catalyst / filter, naturally aspirated / turbocharged, gasoline / diesel engines. The objectives of the test procedures are: To have a robust and consistent approach to assess the performance of exhaust manifold/catalytic converter systems. To optimise the design so that it can achieve the specified design targets. To establish a systematic way to collect and to report data. The use of the CFD-based test procedure for exhaust Product Development (PD) is mandatory since The current test procedure is based on STAR-CD. Page 3

4 Design Variants Close-coupled catalyst for an 1.6L I4 naturally aspirated gasoline application Under floor catalyst for a 2.0L I4 turbocharged gasoline application Close-coupled catalyst for an 3.5L V6 naturally aspirated gasoline application After treatment system for an 2.2L I4 turbocharger diesel application Page 4

5 Benefits of Using CFD for Exhaust PD One key parameter to determine the exhaust after treatment system performance is the amount of precious materials (PGM) used in the catalyst. By combining the use of CFD in exhaust PD and other technology advancements in other areas, such as improved wash coat formulations and calibration techniques, a significant improvement in emissions performance and reduction in PGM cost and weight could be achieved. Stage 4 TWC of 1.6L gasoline engine for B-car with cast exhaust manifold: 1.2L substrate, weighed 10.4 kg Stage 5 TWC of 1.6L gasoline engine for B- and C- car with fabricated exhaust manifold: 1.0L substrate, weighed 5.3 kg Page 5

6 Cost in $ PGM ($) PGM Cost Reduction PGM Cost of a Stage IV TWC after treatment system for a typical 1.6L gasoline engine from 1998 to 2006 Model Year $ $ $ $ $ $ Total PGM Cost at CBP PGM rates Total PGM Cost at April 06 PGM rates $80.00 $79.59 $69.78 $73.90 $60.00 $43.74 $45.06 $49.44 $40.00 $20.00 $17.15 $21.08 $7.78 $21.26 $0.00 C Sigma Job 1 C Sigma 01MY C1 Sigma Job 1 C Sigma 04MY C1 Sigma 04MY C1 Sigma 06MY 1998 MY 2006 MY Courtesy of M. Brogan Page 6

7 Exhaust PGM & Total Costs between European OEMs European OEM 1.6 Petrol St IV Catalyst Internals Estimated Costs Model Year $ $ $ $ $96.89 PGM Cost (CBP rates) Total Cost (CBP rates) $85.04 $80.00 $77.22 $78.37 $75.69 $60.00 $56.68 $47.52 $40.00 $36.73 $37.08 $38.76 $39.51 $24.01 $24.30 $20.00 $7.78 $13.86 $0.00 Ford Focus BMW 116 Audi A3 VW Golf Peugeot 306 Renault Other European OEM Megane Mercedes A150 Vauxhall Astra Courtesy of M. Brogan Page 7

8 Assumptions of the Current Approach Exhaust gas is represented by air. The gas flow in an exhaust system is of a transient nature but the analysis was simplified to a number of steady state analyses. Boundary conditions, such as mass flow rate, are adjusted according to the engine types, e.g. naturally aspirated or turbocharged. Chemical reactions are not included in the simulations. Standard k-epsilon turbulence model with high Y+ for near wall treatment. All wall boundaries are assumed to be adiabatic, e.g. No heat transfer. Substrate of the catalytic converter or filter, e.g. diesel particulate filter, is modelled as porous media. Pressure drop across an uncoated substrate under the specified operation condition is described by the following equation: DP/L = -(av + b)*v a and b are know as permeability coefficients Physical properties of the uncoated substrate are characterised by the open frontal area (OFA), hydraulic diameter (d h ) and material porosity. User subroutines are used to determine the pressure coefficients of the substrates. Page 8

9 Key Features of the Current Procedure The procedure defines (or recommends) certain requirements for performing steady state CFD analysis, such as Software requirements Modelling requirements Mesh requirements and quality Set-up requirements Modelling the substrate Boundary conditions Analysis requirements Post-processing Reporting format Page 9

10 Targets The key design targets (for analytical sign-off) are: Flow Uniformity Index A statistical measure of the gas flow distribution across the catalyst front face. Velocity Index--Location of the high velocity flow and it should be kept away from the edge. Other design parameter: Pressure drop values (system and across the catalyst/filter). Supporting information (reference only): Velocity ratio, space velocity, annular velocity ratio etc. Effects of flow mal-distribution on catalyst front-face Effects of flow mal-distribution on mount durability Page 10

11 Objectives of the Upgrade To upgrade the analytical process from STAR-CD to STAR-CCM+ format. The new process shall maintain all STAR-CD key features, e.g. User subroutine to determine the pressure coefficients Post processing scripts Ease to use As a minimum, the STAR-CCM+ version should replicate most (or ideally all) the things that STAR-CD version can do. Make use of the new modelling techniques, e.g. use Full Momentum instead of Darcy Law for porous material modelling. Using better approaches to determine the convergence. Ideally, the new process should have a minimum impact on the assessment procedure, e.g. use the same design target values. Reduce the turnaround time but maintain quick and high quality analysis. Page 11

12 Current Status Objectives which have been achieved so far: Maintain most of the Prostar/STAR-CD features, key exceptions are 1) use vertex to define value and 2) to calculate the Annular Velocity Ratio. Easy to use, one script for model set-ups etc and one script for analysis/post-processing. Scripts are used to define a large portion of the model set-ups. Applicable to designs with single (turbocharged) or multiple runners (naturally aspirated). Applicable to single and multiple catalyst/filter after treatment systems. Volume meshing (including porous material region) is fully automated. Using field functions to define the pressure coefficients, catalyst (ready), filter (in progress). Unique method to determine the true centre of the catalyst cross-section. Three ways to define the Stopping Criteria. Using field functions to perform the post processing. Scripts to create all the data for reporting. Perform volume meshing/analysis/post-processing in batch mode. The current design target values are applicable. Page 12

13 Work Flow Preparation: Parts need user's input Import the surface model and label the regions with appropritate names Define the options and values for surface re-meshing and volume meshing Define the options and values for boundary conditions and initialization GUI controlled with few user inputs required Run the GUI to define the substrate properties and choose the default model set-ups option Perform the analysis Post-processing Additional model set-ups? YES Modify the model set-ups via CCM+'s (optional) NO Use the results to prepare a summary report for review Page 13

14 GUI Panel Version number Mesh status Number of runners and substrates Options to define the substrate properties (unique ID). One has to repeat this step for cases with multiple substrates Option to use default model set-ups and number of iterations Option to set up pre-swirl at the inlet region Perform analysis with and without post-processing N.B. The GUI panel should be called within the CCM+ Page 14

15 Geometry of the Test Case The geometry is based on a close-coupled diesel oxidation catalyst of an after treatment system for a 2.0L turbo-diesel engine. Page 15

16 Define the Locate Coordinate System Local coordinate system defined by the user in CCM+ will be transferred to the true crosssection centre by using the GUI Page 16

17 STAR-CD vs. STAR-CCM+ ID 1.01 ID 1.06 ID 1.21 ID 1.11 Solver STAR-CD STAR-CCM+ STAR-CCM+ STAR-CCM+ Case ID Mesh (main) Hexa Poly Poly Poly Mesh (substrate) Hexa Poly Poly Poly Diff. scheme First First First Second Turbulence model Standard k-e Standard k-e Realizable k-e Standard k-e Near wall treatment High Y+ High Y+ All Y+ High Y+ Uniformity Index Velocity Index Averaged flow velocity (m/s) System pressure drop (kpa) Pressure drop across the substrate (kpa) Page 17

18 Conclusions In the past several years, Ford used a CFD-based test procedure for catalyst utilization and similar analysis. The use of the such test procedure for design sign-off has been proved very successful. The CFD procedure for catalyst utilization has been upgraded to STAR-CCM+ format. The new procedure has maintained most of the key features as found in the current procedure, such as using physical data to define the pressure coefficients, scripts for post-processing etc. Page 18

19 Thank you for your attention. Any questions? Page 19